Highlights
-
•
Black & Hispanic patients had more incomplete visits preceding & during COVID-19
-
•
Patients from lower COI areas had more incomplete visits before & during COVID-19
-
•
The pandemic did not exacerbate these disparities
-
•
Black patients & patients from lower COI areas had fewer telehealth visits early on
-
•
Institutional, community & legislative advocacy needed to improve access to care
Keywords: access to care, COVID-19 pandemic, pediatric cardiology, race and ethnicity, social determinants of health
Abstract
Objective
To determine racial, ethnic, and socioeconomic status differences in access to ambulatory pediatric cardiology before and during the first year of the COVID-19 pandemic.
Study design
Differences in completion of cardiology clinic visit of patients aged 0-18 years scheduled March 13, 2019, to March 12, 2021, were examined by race, ethnicity, and Childhood Opportunity Index (COI). Visits were categorized as completed, delayed (completed within 6 months), or incomplete (no visit within 6 months). Comparisons were made between prepandemic and pandemic time points (March 2020, pandemic onset; April 2020, volume nadir; July 2020, volume recovery).
Results
Patients who were non-Hispanic Black, Hispanic, and living in very low COI areas had greater incomplete visit rates before and during the pandemic compared with patients who were non-Hispanic White and those living in very high COI areas. Patients who were Hispanic had more incomplete visits during pandemic nadir and recovery compared with prepandemic (P < .001, P = .02). Patients from very low COI areas had more incomplete visits during volume nadir (P = .001). In multivariable analysis, patients from low and very low COI areas had more incomplete visits before the pandemic (aOR 1.65, aOR 1.62); differences resolved during the pandemic. The pandemic altered the disparity for the low COI group only (P = .013).
Conclusions
We saw statistically greater incomplete visit rates for patients who were Hispanic and those from lower COI areas preceding and during the pandemic. The pandemic did not exacerbate these disparities. Efforts to improve access to care are needed for at-risk populations.
Social determinants of health (SDOH), including socioeconomic status (SES), race, ethnicity, and education, have been shown to impact access to quality health care, clinical outcomes, and mortality rates, leading to health inequities in disadvantaged groups.1 In pediatric cardiology, differences in SDOH are associated with worse short- and long-term outcomes, including disparities in treatment, recovery, neurodevelopment, and mortality.2, 3, 4, 5, 6 The COVID-19 pandemic exposed existing disparities in care that were further exacerbated by the pandemic in some populations.
Missed ambulatory care opportunities are emerging as a contributing reason for suboptimal clinical outcomes in children and adults with congenital heart disease (CHD).7, 8, 9 Patients with CHD are more likely to miss cardiology visits and have gaps in care if they are of races or ethnicities that are minoritized, of lower SES, from poorer neighborhoods, have public insurance, or primarily speak a non-English language.10,11 The rapid transition in the US health care system from in-person ambulatory care to telehealth in 2020 in an effort to reduce the risk of COVID-19 transmission may have provided increased access to care for vulnerable patients. However, the effects of this systemic overhaul on existing disparities and pandemic-related complications in accessing ambulatory care are unclear.
In this study, we assessed the sociodemographic distribution of our ambulatory pediatric cardiology population to determine how the COVID-19 pandemic affected racial, ethnic, and SES differences in access to both in-person and telehealth care over the first year of the pandemic. As a measure of SES, we used the Childhood Opportunity Index (COI), a composite score that encompasses many SDOH indicators on the basis of where a family lives. We then compared prepandemic rates of visit completion by race, ethnicity, and COI category at 2 time points during the pandemic: when clinic volume decreased sharply and telehealth acutely accounted for the majority of clinical care and when volume recovered to prepandemic levels by both in-person and telehealth care. We hypothesized that racial and socioeconomic disparities for incomplete visits would heighten with the pandemic. We also hypothesized that telehealth, a health care delivery tool for increasing access during the pandemic, would have lower uptake during the pandemic by families who were minoritized and socially disadvantaged.
Methods
Scheduling Data
The study was conducted in a quaternary pediatric cardiac care center at Boston Children's Hospital, a free-standing children's hospital with multiple satellite ambulatory clinic locations. We included all ambulatory cardiology visits, including in-person and telehealth, scheduled for patients aged 0-18 years, from 1 year before the COVID-19 pandemic through the first year of the pandemic (March 13, 2019, to March 12, 2021). Supplemental ambulatory visits to cardiology, including nutrition consultation, cardiac fitness, and neurodevelopment visits, were excluded.
In anticipation of major changes in our ability to see patients as usual during the early pandemic, the ambulatory leadership team proactively captured patients scheduled for upcoming clinic visits, before restrictions began in the US, working in 3- to 6-month blocks. The record of scheduled visits (named the tracker) served as a template by which all patients were reviewed by their clinical cardiology team and triaged to (1) keep the visit as planned in-person, (2) transition to telehealth, or (3) delay and reschedule the visit. Patients were individually contacted with the recommended care plan in advance of their scheduled appointment. Many patients and families also contacted the clinic directly requesting to delay care early in the pandemic. In addition, referrals decreased as primary care visits were similarly affected. The clinic team outreach, personal family decisions, and reduced referrals all resulted in an acute reduction (∼50%) of scheduled visits at the start of the pandemic, with increase in volume thereafter. In this study, visit data were reviewed retrospectively after the aforementioned changes had occurred and reflected visits that had been triaged and remained scheduled. The tracker was ultimately used for ∼18 months during the pandemic. Regular clinic operations also continued, including the rescheduling of canceled or no-show appointments.
Visits were categorized into 3 groups: “completed” as scheduled, “delayed” completion (completed within 6 months of scheduled date), or “incomplete” (no visit within 6 months of the scheduled date). The 6-month mark was chosen as the likely threshold for returning to re-establish care during the study period based on our data. Although patients could be scheduled for multiple visits within the study period, determination of completion was on the basis of the index visit. Telehealth use before the pandemic was minimal but quickly adapted by ambulatory cardiology providers soon after the pandemic onset.
The primary outcome was the proportion of visits that were completed as scheduled, delayed, or incomplete. The secondary outcome was the proportion of telehealth visits that were completed, including scheduled in-person visits that were converted to telehealth. “Volume nadir,” where typical ambulatory clinic volume plummeted, occurred early in the pandemic (April 2020), and “volume recovery,” where volume returned to near prepandemic levels, occurred later in the pandemic (July 2020, Figure 1), with corresponding control months in April 2019 and July 2019, respectively. This research was determined to be exempt by the Boston Children's Hospital Institutional Review Board (IRB-P00038478).
Figure 1.
Daily in-person and telehealth clinic visits in 2020, before and at the onset of the COVID-19 pandemic (March 2020). Highlighted months correspond to volume nadir (April 2020) and volume recovery (July 2020).
Patient Population
Visit details and patient demographics including race, Hispanic ethnicity, street address, age, sex, insurance, and primary language were retrospectively extracted from the electronic medical record. Race and Hispanic ethnicity were combined into one variable and divided into 5 categories: (1) Non-Hispanic White, (2) Non-Hispanic Black, (3) Hispanic, (4) Asian, and (5) Other. Non-Hispanic ethnicities listed in the electronic medical record were not incorporated into our race and ethnicity demographic. To assess neighborhood-level SDOH, we used the COI 2.0, a validated, census-tract composite index that integrates 29 weighted indicators across 3 SDOH domains: education, health and environment, and social and economic.12 COI z-scores are nationally normalized and stratified into 5 levels—very low, low, moderate, high, and very high—each representing 20% of the US child population. Each patient's residential address from 2021 was geocoded, mapped to the latest US census tract available (2020, 2015, or 2010), and assigned a corresponding COI level using the COI 2.0 database for the latest available year (2012 or 2017). Patients with missing or non-US addresses were excluded from the COI analysis.
Statistical Analyses
Descriptive statistics were reported for overall data. Visit and sociodemographic comparisons were made using χ2 tests. To assess whether the pandemic affected visit completion rates differently on the basis of race, ethnicity, or COI, generalized linear models with a Poisson link function were used that included a binary pandemic variable (pre- vs during), COI level (reference category: very high), or race (non-Hispanic White vs all other categories), as well as their interaction (eg, COI∗pandemic). Multivariable logistic regression models were performed using a similar approach by adjusting for covariates that were deemed unlikely confounders of race, ethnicity, or COI: age at visit, sex, and primary language. Given the complex interconnections between race, ethnicity, and COI,13 we chose not to adjust for COI during the multivariate analysis for race and ethnicity, and vice versa.
Results
Clinic Volume and Patient Characteristics
Before the COVID-19 pandemic, our cardiology clinic averaged approximately 19 000 visits annually (Figure 1). At the start of the pandemic in the US (April 2020), clinic volumes decreased sharply by 57%, representing the lowest point during the pandemic. However, clinic volumes had recovered to prepandemic levels by July 2020. Over the course of the first year of the pandemic, total clinic volume decreased by 4.5% compared with the year prior. Before the pandemic, only 0.8% of cardiology visits were conducted via telehealth. In response to the pandemic, telehealth was rapidly adopted and accounted for 46% of all visits during the first year of the pandemic.
A total of 16 569 unique patients completed visits either in the year before or during the first year of the pandemic, with 85% residing in Massachusetts. Table I summarizes the breakdown of completed, delayed, and incomplete visits during the pandemic nadir and recovery periods compared with prepandemic levels. Both delayed and incomplete visit rates were significantly greater during the pandemic when volume recovered, compared with the prepandemic period (July 2020 vs 2019, 23% vs 17% and 15% vs 12%, respectively; P < .001).
Table I.
Proportion of scheduled clinic visits that were completed, delayed, or incomplete during pandemic volume nadir (April 2020) and volume recovery (July 2020), compared with the same months in the prepandemic year
| Outcomes of scheduled visits | Prepandemic (2019), No. (%) | Pandemic (2020), No. (%) | P value |
|---|---|---|---|
| April | 2222 | 1709 | <.001 |
| Completed as scheduled | 1535 (69.1) | 666 (39.0) | |
| Delayed completion within 6 months | 391 (17.6) | 699 (40.9) | |
| Incomplete by 6 months | 296 (13.3) | 344 (20.1) | |
| July | 2409 | 2714 | <.001 |
| Completed as scheduled | 1708 (70.9) | 1688 (62.2) | |
| Delayed completion within 6 months | 406 (16.9) | 623 (23.0) | |
| Incomplete by 6 months | 295 (12.3) | 403 (14.9) |
Patient characteristics from the year before and the first year of the pandemic are shown in Table II. In the first pandemic year, a greater percentage of infants and adolescents were seen compared with the year prior. Completed appointments shifted toward patients who were non-Hispanic White during the pandemic compared with all other races and ethnicities combined (75% vs 71%, P < .001). Of note, nearly one-quarter of patients lacked documented race and ethnicity in the electronic medical record; these patients were excluded from further analyses. At our institution, race and ethnicity are asked by the front desk staff at the time of visit check-in, if not completed during self check-in. Collection of this demographic decreased during the pandemic when there was more reliance on telehealth which lacked an organized check-in process.
Table II.
Characteristics of unique patients with completed visits 1 year before and during the first COVID-19 pandemic year
| Patient characteristics | Prepandemic n = 14 257 |
Pandemic n = 7701 |
P value |
|---|---|---|---|
| No. (%) | No. (%) | ||
| Female | 6842 (48.0) | 3870 (50.3) | .001 |
| Age at index visit in years, mean (SD) | 8.79 (6.01) | 8.86 (6.26) | .40 |
| Age at index visit in years | <.001 | ||
| <1 | 1767 (12.4) | 1294 (16.8) | |
| 1-4 | 2757 (19.3) | 1210 (15.7) | |
| 5-12 | 4691 (32.9) | 2279 (29.6) | |
| 13-18 | 5042 (35.4) | 2918 (37.9) | |
| Race and ethnicity | <.001 | ||
| Non-Hispanic White | 8119 (71.0) | 4226 (74.8) | |
| Non-Hispanic Black | 759 (6.6) | 351 (6.2) | |
| Hispanic | 1275 (11.2) | 561 (9.9) | |
| Asian | 512 (4.5) | 197 (3.5) | |
| Other | 767 (6.7) | 317 (5.6) | |
| Missing race and ethnicity data | 2825 (19.8) | 2049 (26.6) | <.001 |
| Non-English primary language (missing data: n = 994) | 1504 (11.1) | 649 (8.7) | <.001 |
| Public insurance (missing data: n = 84) | 4268 (30.1) | 2286 (29.8) | .71 |
| COI | .18 | ||
| Very low | 1893 (13.8) | 977 (13.0) | |
| Low | 2255 (16.5) | 1185 (15.8) | |
| Middle | 2678 (19.6) | 1530 (20.4) | |
| High | 3120 (22.8) | 1727 (23.0) | |
| Very high | 3728 (27.3) | 2092 (27.9) | |
| Missing COI data | 583 (4.1) | 190 (2.5) | <.001 |
The proportion of patients with a non-English primary language decreased from 11% to 9% during the pandemic, whereas the proportion with public insurance remained stable at 30%. Approximately one-half of all patients lived in census tracts classified as high or very high COI, with no significant change in COI distribution during the pandemic. Figure 2 shows COI categories by race and ethnicity. Patients who were non-Hispanic White and Asian were more likely to live in areas with greater COI, whereas patients who were non-Hispanic Black and Hispanic were more likely to live in areas with low or very low COI (P < .001).
Figure 2.
Relationship between COI and race and ethnicity for unique patients, including before and during the first pandemic year (n = 16 569). P values were significant (<.001) for non-Hispanic White vs all other categories.
Incomplete Visits by Race, Ethnicity, and COI
Incomplete visit rates were greater among patients who were non-Hispanic Black and Hispanic compared with patients who were non-Hispanic White across all time points both before and during the pandemic (Figure 3A), only reaching statistical significance in patients who were Hispanic. During the peak disruption in clinic operations (April 2020) and the early recovery phase (July 2020), patients who were Hispanic experienced a disproportionately greater rate of incomplete visits compared with the prepandemic period (29% vs 16%, P < .001, and 20% vs 12.4%, P = .02, respectively). Incomplete visit rates also varied by COI. Patients living in areas with very low COI had greater incomplete visit rates at every time point compared with those from very high COI areas (Figure 3B); this disparity was particularly pronounced early in the pandemic, with significantly more incomplete visits among patients from very low COI areas (29% vs 15%, April 2020 vs 2019, P = .001).
Figure 3.
Incomplete visits before and during the pandemic stratified by A, race and ethnicity, and B, COI. Comparison made with non-Hispanic White or very high COI, respectively.
In unadjusted analyses, comparing July in the prepandemic year to the first pandemic year, only patients who were Hispanic had significantly greater odds of incomplete visits during the pandemic (OR 1.57, 95% CI 1.09-2.25; Table III). However, after we adjusted for age at visit, sex, and primary language, this association was no longer statistically significant. Moreover, all interaction terms between race and ethnicity and the pandemic period (race/ethnicity∗pandemic) had P values > .1, suggesting that the pandemic did not significantly alter existing racial and ethnic disparities in incomplete visits.
Table III.
Multivariable analysis of rates of incomplete visits by race, ethnicity, and COI
| Patient characteristic | aOR (95% CI) |
|
|---|---|---|
| July 2019 | July 2020 | |
| Total scheduled visits, No. | 2409 | 2714 |
| Incomplete visits, No. (%) | 295 (12.3%) | 403 (14.9%) |
| Multivariable models for race and ethnicity | ||
| Unadjusted model (incomplete/total) | 232/1957 | 307/2145 |
| Non-Hispanic White | REF | REF |
| Non-Hispanic Black | 1.06 (0.61, 1.84) | 1.06 (0.64, 1.73) |
| Hispanic | 1.09 (0.70, 1.69) | 1.57 (1.09, 2.25)∗ |
| Asian | 1.03 (0.55, 1.92) | 1.29 (0.74, 2.25) |
| Other | 1.27 (0.78, 2.05) | 1.09 (0.67, 1.79) |
| Adjusted model: adjusted for age at visit, sex, language (incomplete/total) | 220/1852 | 296/2056 |
| Non-Hispanic White | REF | REF |
| Non-Hispanic Black | 1.16 (0.67, 2.02) | 1.13 (0.68, 1.86) |
| Hispanic | 1.26 (0.78, 2.04) | 1.43 (0.93, 2.21) |
| Asian | 1.11 (0.57, 2.14) | 1.24 (0.69, 2.22) |
| Other | 1.45 (0.85, 2.47) | 1.12 (0.67, 1.87) |
| Multivariable models for COI | ||
| Unadjusted model (incomplete/total) | 277/2263 | 384/2639 |
| Very high COI | REF | REF |
| High COI | 1.15 (0.79, 1.68) | 1.13 (0.85, 1.52) |
| Middle COI | 1.03 (0.69, 1.53) | 0.80 (0.57, 1.11) |
| Low COI | 1.62 (1.11, 2.37)∗ | 0.82 (0.58, 1.16) |
| Very low COI | 1.49 (0.99, 2.22) | 1.19 (0.83, 1.69) |
| Adjusted model: adjusted for age at visit, sex, language (incomplete/total) | 261/2141 | 369/2535 |
| Very high COI | REF | REF |
| High COI | 1.19 (0.81, 1.76) | 1.14 (0.85, 1.54) |
| Middle COI | 1.07 (0.71, 1.61) | 0.77 (0.55, 1.09) |
| Low COI | 1.65 (1.11, 2.45)∗ | 0.79 (0.55, 1.13) |
| Very low COI | 1.62 (1.05, 2.51)∗ | 1.15 (0.78, 1.71) |
REF, reference category.
Denotes significant aOR.
Regarding COI, unadjusted models revealed that patients living in low COI areas before the pandemic had significantly greater odds of incomplete visits compared with those living in very high COI areas (OR 1.62, 95% CI 1.11-2.37; Table III). Although patients from very low COI areas also showed elevated odds (OR 1.49), the 95% CI narrowly included 1 (0.99-2.22). No significant OR was seen during the pandemic. In the adjusted model, patients from both low and very low COI areas remained significantly more likely to have incomplete visits pre-pandemic (aOR 1.65, 95% CI 1.11-2.45, and aOR 1.62, 95% CI 1.05-2.51, respectively). Only the interaction between low COI and the pandemic period (low COI∗pandemic) reached statistical significance (P = .013), indicating that the pandemic altered the pattern of disparities for this group from worse to better. All other COI∗pandemic interaction terms were not significant (P > .1).
Telehealth Use by Race, Ethnicity, and COI
Differences in completed telehealth visits by race and ethnicity during volume nadir (April 2020) and recovery (July 2020) are shown in Figure 4A. At the height of clinic disruption in April 2020, patients who identified as non-Hispanic Black were significantly less likely to participate in telehealth visits compared with patients who identified as non-Hispanic White (70% vs 89%, P = .002). However, by July 2020, there was no statistically significant difference in telehealth completed visit rates by race or ethnicity. Differences in telehealth use were also observed by COI. While telehealth access improved for most groups by July 2020, patients living in very low COI areas continued to have significantly lower telehealth use rates (P < .001; Figure 4B), indicating a persistent disparity.
Figure 4.
Difference in completed telehealth visits during the COVID-19 pandemic by A, race and ethnicity, and B, COI. Comparison made to non-Hispanic White or very high COI, respectively.
Discussion
In the present study, we investigated how the COVID-19 pandemic impacted access to ambulatory cardiology care for children presenting to a single, large academic center providing care for the New England region. In our cohort, we found that incomplete visits occurred at greater rates for patients who were non-Hispanic Black and Hispanic compared with patients who were non-Hispanic White, preceding and throughout the first pandemic year, although only patients who were Hispanic had statistically significantly increased rates of incomplete visits during the pandemic. Compared with children living in areas with very high COI, those residing in low COI areas had significantly greater rates of incomplete visits in the early pandemic. Families of non-Hispanic Black and Hispanic race in our study were disproportionately from areas of very low and low COI, congruent with correlations shown elsewhere.14,15
Although racial, ethnic, and COI differences in visit completion rates were more pronounced during the pandemic, the pandemic did not statistically alter disparities for any group except those living in low COI areas. Notably, in adjusted multivariable models, the disparities seen prepandemic for patients living in low and very low COI areas were not present during the pandemic. Regardless, the prepandemic disparities indicate that further policy work is needed to achieve health equity within the pediatric ambulatory cardiology settings. Our research aligns with previous studies across multiple ambulatory pediatric subspecialties, highlighting the racial and socioeconomic disparities in access that predated and persisted through the pandemic.16,17
Similar to other pediatric cardiology centers, experience with telehealth in our ambulatory cardiology clinic was limited before the COVID-19 pandemic, with a rapid uptake of broad telehealth services soon after the pandemic onset.18,19 Nearly all of our ambulatory providers were trained in the use of telehealth within 2 weeks of the pandemic onset. Clinic volume over the first pandemic year relied heavily on telehealth, with almost one-half of all visits completed virtually vs less than 1% of visits before the pandemic. Telehealth emerged to increase access to care safely by avoiding increased risk of infection transmission, plus it had the potential to relieve the burdens of transportation cost and lost wages from extended time off work.
Yet, the benefits of telehealth on disparities influencing access to care remain controversial. In our study, there were proportionately fewer completed telehealth visits by families of non-Hispanic Black race and by families living in areas with very low COI compared with families of non-Hispanic White race and families living in very high COI areas. Studies in different pediatric populations have shown conflicting findings related to racial and ethnic disparities in telehealth use during the COVID-19 pandemic, from decreased uptake in families of Black race20 and Hispanic ethnicity,18,21 to increased use in families of Hispanic ethnicity,22,23 to no association.24 The low uptake of telehealth throughout the early pandemic for our patients living in very low COI areas was echoed in some reports yet contradicted in others.21,24 To understand whether the differences in telehealth we observed in this study persisted, we reviewed our clinic's telehealth use by race and ethnicity during fiscal year 2025. Telehealth continues to be used less by patients who are non-Hispanic Black compared with patients who are non-Hispanic White, without significant differences among other racial and ethnic groups, indicating that ongoing work is still needed. Data were not available on the basis of COI.
When comparing the demographics of unique patients with completed visits in the year before and the first year of the COVID-19 pandemic, we saw a significant decrease during the pandemic in the proportion of families who primarily spoke non-English languages, which aligns with previous findings that noted a significantly lower uptake of telehealth among patients with limited English language proficiency in pediatric subspecialty clinics and adult clinics.18,25,26 Surprisingly, we did not see a decrease in proportion during the pandemic for those with public insurance, contrary to trends seen in other studies.21,24 Given 85% of our study population lived in Massachusetts, this difference may be due to the expansive protections of Massachusetts public insurance (MassHealth) relative to other states’ plans, such as patients remaining insured for 12 continuous months even with family income change, which may have mitigated the effect of the pandemic by insurance status in Massachusetts compared with other states. In addition, we saw a greater proportion of adolescents during the pandemic, despite adolescents being at high risk of becoming lost to care.11,27 The pandemic raised great concern of worse outcomes for patients with underlying cardiac disease, perhaps fostering improved follow-up of adolescents who receive more intermittent cardiac care.
Disparities in SDOH during COVID-19
As we have shown, the COVID-19 pandemic brought to light disparities in health care access and use for patients who are minoritized and those of lower SES, likely related to multifactorial reasons. Appointments may be intentionally or unintentionally missed due to a number of competing priorities, such as taking time off work, finding childcare for other children, commuting time and cost, misunderstanding the need for follow-up, and facing complicated processes to cancel or reschedule appointments. The pandemic compounded these situations by adding heightened fears of contracting COVID-19 while balancing virtual school for children, navigating remote work at home, and exacerbating already limited financial resources for some vulnerable families.
Other potential pandemic-related factors limiting access to care included restrictions on the number of people who could accompany a patient to a visit, which placed an additional burden of coordination on families of multiple children. Medical centers also rescheduled large numbers of visits while rapidly transitioning to a telehealth platform, increasing the risk of patients being lost to care. Although health care teams received some training on using telehealth, patients and families received none. Challenges with using telehealth for families with fewer resources or living in areas with lower COI could be related to the need for reliable broadband internet connection, limited access to working devices and software, poor digital literacy, or discomfort with the lack of in-person examination and discussion.
Policy Implications
The described disparities in ambulatory pediatric care predate the pandemic and require efforts on all levels—institutional, community, and legislative—to improve access particularly for minoritized and socially disadvantaged families. On an institutional level, universal screening for health-related social needs can identify families at risk and provide an opportunity for wrap-around care. Urgent and chronic needs can also be addressed through social work and patient navigator resources, particularly if language and literacy barriers are present.
At our center, in follow up to our study findings, we have used population health strategies for broad and proactive identification of patients with missed or delayed care for whom robust outreach programs may facilitate re-engagement with the medical system. Our institution has implemented a targeted working group to ensure patients receive appropriate follow-up, particularly for those with moderate-to-severe CHD, who are at risk of worse outcomes from gaps in care. Patients who have previously missed appointments or sought care in satellite clinics with low show rates are proactively contacted in advance of their next appointments. In addition, we are expanding SDOH screening across all of our ambulatory clinics as well as during patient admissions in anticipation of greater support at time of discharge. Interviewing vulnerable families who have missed care during the pandemic has informed our strategies to facilitate ambulatory access. Lastly, we are bolstering patient, family, and provider educational materials around high-risk periods of becoming lost to care, such as during the transition to adult care, with insight from our Family Advisory Council.
However, vulnerable families require additional support beyond the hospital to address external pressures influencing access to care, such as demanding work schedules or childcare needs. Vouchers or gift cards for transportation, parking, gas, and meals can reduce the risk of missed care opportunities due to financial constraints. On a community level, care providers can facilitate connection to community programs that reduce food or housing insecurity. On a policy level, advocacy for telehealth visits provides opportunities for families whose burdens are eased by remote care. Government-funded insurance with expansive protections (such as MassHealth) is critical in supporting vulnerable children and those transitioning to adulthood with CHD and other chronic diseases to ensure access to longitudinal coverage of care.
These are only a few select recommendations out of the many changes needed. Reducing disparities in ambulatory care requires constant reassessment of clinic operations, flexible care opportunities, resources to address SDOH, and tireless effort and advocacy. By coordinating outreach across multiple levels, care centers can more effectively address the systemic effects of SDOH inequity to reduce existing disparities in ambulatory pediatric care.
Limitations
Our study has limitations worth noting. Our results represent data from a single, high-volume quaternary pediatric cardiac care center, which may not be applicable to other institutions or parts of the country. Despite great efforts to triage patients originally scheduled before the pandemic, there were large fluctuations in clinic volume in the early pandemic with an acute drop of approximately 50%. Our findings are based on patients who were scheduled for appointments, and do not describe patients unscheduled for planned delays in care or those who may be lost to care. In addition, race and ethnicity designation were missing or unspecified in ∼25% of patients and increased with the transition to telehealth as there was no check-in process where the information was ordinarily collected. Addresses abstracted from the electronic medical record may differ over time, leading to inaccurate COI analyses. Also, housing insecurity worsened during the pandemic, forcing some families to move locations and potentially affecting address data. We used COI 2.0, since the most current COI 3.0, which includes new indicators (44 compared with 29 in COI 2.0) for years 2012-2023, was published after our data analysis was complete. It is possible that COI 2.0, which draws data from years 2012 or 2017, did not accurately reflect neighborhood-level SDOH during our study period 2019-2021. While the COI measure incorporates many community-level SDOH, the complex interplay among race, ethnicity, and COI related to disparities in access to care were not elucidated in this study and require further investigation.
Conclusions
Racial, ethnic, and SES disparities exist in ambulatory pediatric cardiology. Although the COVID-19 pandemic brought these disparities to light, the magnitude of the disparities was not worsened by the pandemic. Efforts on all levels are needed to reduce disparities in care and improve health equity.
CRediT authorship contribution statement
Larissa Wenren: Writing – review & editing, Writing – original draft, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Emily Bucholz: Writing – review & editing, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Maria Perez Carballo: Writing – review & editing, Methodology, Investigation, Conceptualization. Francis Fynn-Thompson: Writing – review & editing, Methodology, Investigation, Conceptualization. Kathy Jenkins: Writing – review & editing, Methodology, Investigation, Conceptualization. Jean Connor: Writing – review & editing, Methodology, Investigation, Conceptualization. Sarah de Ferranti: Writing – review & editing, Visualization, Validation, Methodology, Investigation, Conceptualization. Susan F. Saleeb: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Methodology, Investigation, Data curation, Conceptualization.
Declaration of Competing Interest
This work was supported by the Boston Children's HospitalFred Lovejoy Housestaff Research and Education Fund. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Glossary
Social determinants of health (SDOH): The nonmedical factors that influence health outcomes. Conditions in which people are born, grow, work, live, worship, and age, and include a wide set of forces and systems that shape daily life such as economic policies and systems, development agendas, social norms, social policies, and political systems, as well as the effects of centuries of racism.1
Childhood Opportunity Index (COI): A validated, census-tract composite index that integrates 29 weighted indicators across 3 SDOH domains: education, health and environment, and social and economic.12
Race: A social construct used to group people often based on physical appearance and social factors.
Ethnicity: A social construct used to group people who share a common cultural background or descent.
References
- 1.Centers for Disease Control and Prevention Social Determinants of Health (SDOH) https://www.cdc.gov/about/priorities/why-is-addressing-sdoh-important
- 2.Best K.E., Vieira R., Glinianaia S.V., Rankin J. Socio-economic inequalities in mortality in children with congenital heart disease: a systematic review and meta-analysis. Paediatr Perinat Epidemiol. 2019;33:291–309. doi: 10.1111/ppe.12564. [DOI] [PubMed] [Google Scholar]
- 3.Bucholz E.M., Sleeper L.A., Goldberg C.S., Pasquali S.K., Anderson B.R., Gaynor J.W., et al. Socioeconomic status and long-term outcomes in single ventricle heart disease. Pediatrics. 2020;146 doi: 10.1542/peds.2020-1240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dionne A., Bucholz E.M., Gauvreau K., Gould P., Son M.B.F., Baker A.L., et al. Impact of socioeconomic status on outcomes of patients with Kawasaki disease. J Pediatr. 2019;212:87–92. doi: 10.1016/j.jpeds.2019.05.024. [DOI] [PubMed] [Google Scholar]
- 5.Shi Y., de Groh M., Bancej C. Socioeconomic gradients in cardiovascular risk in Canadian children and adolescents. Health Promot Chronic Dis Prev Can. 2016;36:21–31. doi: 10.24095/hpcdp.36.2.02. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tjoeng Y.L., Jenkins K., Deen J.F., Chan T. Association between race/ethnicity, illness severity, and mortality in children undergoing cardiac surgery. J Thorac Cardiovasc Surg. 2020;160:1570–1579.e1. doi: 10.1016/j.jtcvs.2020.06.015. [DOI] [PubMed] [Google Scholar]
- 7.Awh K., Venuti M.A., Gleason L.P., Rogers R., Denduluri S., Kim Y.Y. Clinic nonattendance is associated with increased emergency department visits in adults with congenital heart disease. Congenit Heart Dis. 2019;14:726–734. doi: 10.1111/chd.12784. [DOI] [PubMed] [Google Scholar]
- 8.Diller G.P., Orwat S., Lammers A.E., Radke R.M., De-Torres-Alba F., Schmidt R., et al. Lack of specialist care is associated with increased morbidity and mortality in adult congenital heart disease: a population-based study. Eur Heart J. 2021;42:4241–4248. doi: 10.1093/eurheartj/ehab422. [DOI] [PubMed] [Google Scholar]
- 9.Rosamilia M.B., Williams J., Bair C.A., Mulder H., Chiswell K.E., D'Ottavio A.A., et al. Risk factors and outcomes associated with gaps in care in children with congenital heart disease. Pediatr Cardiol. 2024;45:976–985. doi: 10.1007/s00246-024-03414-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Demianczyk A.C., Behere S.P., Thacker D., Noeder M., Delaplane E.A., Pizarro C., et al. Social risk factors impact hospital readmission and outpatient appointment adherence for children with congenital heart disease. J Pediatr. 2019;205:35–40.e1. doi: 10.1016/j.jpeds.2018.09.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zaidi A.H., Saleeb S.F., Gurvitz M., Bucholz E., Gauvreau K., Jenkins K.J., et al. Social determinants of health including child opportunity index leading to gaps in care for patients with significant congenital heart disease. J Am Heart Assoc. 2024;13 doi: 10.1161/JAHA.122.028883. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Noelke C., McArdle N., Baek M., Huntington N., Huber R., Hardy E., et al. Child opportunity index 2.0. Technical documentation. diversitydatakids.org. 2020. https://www.diversitydatakids.org/sites/default/files/2020-02/ddk_coi2.0_technical_documentation_20200212.pdf [DOI] [PubMed]
- 13.Duong S.Q., Elfituri M.O., Zaniletti I., Ressler R.W., Noelke C., Gelb B.D., et al. Neighborhood childhood opportunity, race/ethnicity, and surgical outcomes in children with congenital heart disease. J Am Coll Cardiol. 2023;82:801–813. doi: 10.1016/j.jacc.2023.05.069. [DOI] [PubMed] [Google Scholar]
- 14.diversitydatakids.org Child Opportunity Index 2.0 database: racial/ethnic patterns of child opportunity. 2023. https://www.diversitydatakids.org/research-library/data-visualization/racialethnic-patterns-child-opportunity
- 15.Ioannidis J.P.A., Powe N.R., Yancy C. Recalibrating the use of race in medical research. JAMA. 2021;325:623–624. doi: 10.1001/jama.2021.0003. [DOI] [PubMed] [Google Scholar]
- 16.Jones M.K., O'Connell N.S., Skelton J.A., Halvorson E.E. Patient characteristics associated with missed appointments in pediatric subspecialty clinics. J Healthc Qual. 2022;44:230–239. doi: 10.1097/JHQ.0000000000000341. [DOI] [PubMed] [Google Scholar]
- 17.Phan T.T., Enlow P.T., Lewis A.M., Arasteh K., Hildenbrand A.K., Price J., et al. Persistent disparities in pediatric health care engagement during the COVID-19 pandemic. Public Health Rep. 2023;138:633–644. doi: 10.1177/00333549231163527. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cahan E.M., Maturi J., Bailey P., Fernandes S., Addala A., Kibrom S., et al. The impact of telehealth adoption during COVID-19 pandemic on patterns of pediatric subspecialty care utilization. Acad Pediatr. 2022;22:1375–1383. doi: 10.1016/j.acap.2022.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Uscher-Pines L., McCullough C., Dworsky M.S., Sousa J., Predmore Z., Ray K., et al. Use of telehealth across pediatric subspecialties before and during the COVID-19 pandemic. JAMA Netw Open. 2022;5 doi: 10.1001/jamanetworkopen.2022.4759. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Katzow M.W., Steinway C., Zuzarte A., Chen J., Fishbein J., Jan S. Sociodemographic disparities in ambulatory pediatric telemedicine utilization during COVID-19. Telemed J E Health. 2024;30:57–66. doi: 10.1089/tmj.2023.0005. [DOI] [PubMed] [Google Scholar]
- 21.George P.E., Kc D., Greenleaf M., Shah J., Lam W.A., Hawkins C.M. Bridging the divide: unintended consequences of the shift to home-based telemedicine. J Pediatr. 2024;269 doi: 10.1016/j.jpeds.2023.113719. [DOI] [PubMed] [Google Scholar]
- 22.Baker-Smith C.M., Sood E., Prospero C., Zadokar V., Srivastava S. Impact of social determinants and digital literacy on telehealth acceptance for pediatric cardiology care delivery during the early phase of the COVID-19 pandemic. J Pediatr. 2021;237:115–124.e2. doi: 10.1016/j.jpeds.2021.06.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Barney A., Mendez-Contreras S., Hills N.K., Buckelew S.M., Raymond-Flesch M. Telemedicine in an adolescent and young adult medicine clinic: a mixed methods study. BMC Health Serv Res. 2023;23:680. doi: 10.1186/s12913-023-09634-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Vaughan R.M., Moore J.A., Moreno J.S., Dyer K.J., Oluyomi A.O., Lopez K.N. Remote care adoption in underserved congenital heart disease patients during the COVID-19 era. Pediatr Cardiol. 2023;44:404–412. doi: 10.1007/s00246-022-03042-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Eberly L.A., Kallan M.J., Julien H.M., Haynes N., Khatana S.A.M., Nathan A.S., et al. Patient characteristics associated with telemedicine access for primary and specialty ambulatory care during the COVID-19 pandemic. JAMA Netw Open. 2020;3 doi: 10.1001/jamanetworkopen.2020.31640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kim P.C., Tan L.F., Kreston J., Shariatmadari H., Keyoung E.S., Shen J.J., et al. Socioeconomic factors associated with use of telehealth services in outpatient care settings during the COVID-19. BMC Health Serv Res. 2024;24:446. doi: 10.1186/s12913-024-10797-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Moore J.A., Sheth S.S., Lam W.W., Alexander A.J., Shabosky J.C., Espaillat A., et al. Hope is no plan: uncovering actively missing transition-aged youth with congenital heart disease. Pediatr Cardiol. 2022;43:1046–1053. doi: 10.1007/s00246-022-02823-1. [DOI] [PMC free article] [PubMed] [Google Scholar]